Dailydave: OWASP AppSec 2004

[dave <dave () immunitysec com>]

So I wanted to report on the OWASP conference last weekend after they 
put the slides up, but I  haven't seen that happen yet. So I'll just go 
over it and pretend you can see them.
First of all, for a Saturday and Sunday in Hoboken - the thing thing was 
packed. People came from all over the place. Mostly NYC and DC, but also 
a fair number from London and other places. I've noticed a trend towards 
holding conferences in academic settings. It makes a lot of sense 
because a tiny room at a hotel is 1K a day. This means a big room at a 
hotel, which gets you nothing special, is a lot of money. So you have to 
make the tickets really pricey. I didn't pay for OWASP, since I was 
speaking, but 300 dollars is not that much.
Anyways, on to the talks. I thought the general tone of the conference 
was very much like an open source project's conference, and not like 
BlackHat, Defcon, PacSec, or any of the security cons I've been to. It 
was interesting to see how much support OWASP has from various banks and 
large financials here in the city, which all want OWASP to succeed so 
they can save money on their bottom line, much like they want Linux to 
succeed.
The official topic was "cutting through vendor hype", although I think 
few talks addressed that directly. Most of the talks were high level, 
with Dinis Cruz's being one of the rare ones that was not. Apparantly 
people are running their ASP.NET hosting environments in "Full Trust" 
which means that they rely on having separate user ID's for security, 
which they can't do for some reason. What was also interesting was that 
few people in the room had any knoweldge of Win32. Windows really hasn't 
penetrated this market at all. Dinis noticed that there are SYSTEM 
identity tokens popping into random processes in Windows 2000. I told 
them about that 2 years ago while reviewing IIS 6, and we determined it 
was harmless, but Dinis thinks maybe if you use DuplicateToken on it 
(it'll return False, but still work), you can do something. I never 
figured it out.
I challenged all the "static analysis" software vendors in the room (at 
least one came up and answered) to analyze pre-TESO-fixes CVS and tell 
me if they found any of the bugs. One of them did say they would do some 
open source projects and get back to me, and if they do, you will read 
about it here, and if they don't you should not buy their tool (or any 
static analysis tool that has not proven itself in this way).
One thing I liked was the new oPORTAL stuff, or whatever they're calling 
their pet portal project. A portal is different from a content 
management system, it turns out. But in any case, it looks like it'll be 
nice, when they get it releasable.
WebGoat, on the other hand, is fucking fantastic. It's a breeze to 
install (one click), fully configurable, and provices point and click 
"lessons" which teach people how to look at web security. The only thing 
missing is more lessons. (esp. advanced lessons).
The only talk I had trouble in was *"Security Considerations in the 
System Development Life Cycle of Web Services-based Systems - Toto, 
We're Not in Kansas Any More . . .* - George Capehart, Founding Member 
of Capehart Associates LLC" He's from the south, so he talked really 
slow, and the talk started out really interesting, but I lost it with 
the slow pace. Also, each slide was 10000 words crammed into the page, 
so it was impossible to really read at a distace. Slides should have 
very few words. Actually, the Japanese do totally crazy stuff with 
slideshows. You have to see it to believe it. It's an experience. I tend 
to stick with the lists of bullets and pictures. But that would not fly 
in Japan. It would look half-hearted, like I whipped it up this morning.
There were a lot of people stabbing at that holy grail - meaningful 
metrics. It's an impossible thing. One person's "High risk" is another 
person's "Low risk" and that's always the way it's going to be.
That's all I have. If anyone has more comments on it, feel free to send 
them on in. Overall, I think it was a huge success - people networked, 
people learned, etc. I learned at least one thing, which is my criteria 
for success.
-dave


_______________________________________________
Dailydave mailing list
Dailydave () lists immunitysec com
http://www.immunitysec.com/mailman/listinfo/dailydave